I'm not sure what is the example supposed to prove.

That what I said doesn't always hold?

Of course it doesn't.

That's why I wrote about possibilities, and only said that the prior that the person preaching X doesn't follow X "somewhat lessens" the possibility that X is true.

I never said that someone "preaching X and doing the opposite" proves that X is not true.

If I had said that, your example would be a valid refutation. But I haven't, and it's only attacking a straw man.

Besides, the only reason that your example case seems obviously wrong to you is because you're taking as a given what you are supposed: you already know that smoking is bad.

If you didn't, and someone said to you "smoking is bad for you", while he himself was smoking, you would be justifiably suspicious to consider whether it's not actually bad, and the other person is lying or hiding something from you (e.g. to keep all the cigarettes for himself).

My point is that this thinking is a fallacy precisely because it is orthogonal to the argument being made. It is neither evidence for nor against the argument. You are making some sort of argument for heuristics, but heuristics are called such precisely because they are not solutions.

The correct answer is to independently validate the claims, or evaluate the merits of an argument on their own ground. That's tedious and often not worth the effort, and that's OK. We use heuristics all the time -- you don't ask your friend for his proof, because you use the heuristic of trust instead. But at least acknowledge that it's a heuristic and stop trying to claim it as a solution.

Thus proving once and for all that smoking isn't bad for you!